Synopsis The left main landing gear of the Fokker F28-Mk1000 began to shimmy immediately after touchdown when landing at Calgary. Brakes were applied to slow the aircraft in an attempt to control the shimmy, but the oscillations continued until both left main wheels and brake assemblies separated from the axles. After the aircraft came to a stop, the passengers and crew were evacuated without incident through the forward main cabin door. Site examination revealed that the upper torque link failed within the first 200 feet of the landing roll, and the wheels separated about 1,450 feet from touchdown. There was substantial damage to the oleo lower sliding member, wheels, tires, brakes, and left inboard and outboard flaps. The crew reported that the lift dumpers (spoilers) had been armed on the approach. They were found to be deployed when the aircraft came to a stop, but the crew was uncertain as to when they had deployed. Other Factual Information The aircraft had recently been out of service because of a shimmy problem. The problem had been isolated to the right main gear and was rectified by the replacement of an out-of-balance inboard wheel and tire. A successful flight test and positioning flight had preceded the occurrence flight when the left main gear developed a shimmy. In normal operation, any torsional loads on the lower oleo sliding member from the wheel assemblies on the dual axles are transferred through the torque links (T/Ls) to the stationary upper oleo member. The T/Ls prevent rotation about the vertical axis to maintain wheel alignment in the fore-and-aft direction throughout the vertical range of oleo travel. In maintaining proper wheel alignment, the T/Ls stabilize any torsional oscillations (shimmy). By their design, T/Ls have their maximum mechanical advantage and control when the oleo is compressed, and minimum advantage when the oleo is fully extended. Both left and right oleo assemblies were disassembled and examined in the operator's overhaul (O/H) shop. Dimensions were generally found to be within service limits, with the exception of the fit of the T/L pins to the main oleo lug bushings and the sliding member lug bushings. The clearances in the main bushings were from 0.0012 inches to 0.0034 inches, and in the sliding member bushings were from 0.0021 inches to 0.0154 inches. The manufacturer's O/H Manual specifies an interference fit, whereby the pin is pressed into the bushings without any clearance. This interference fit is necessary to minimize free play and ensure stability of the moving parts of the landing gear -- that is, to prevent shimmy. The pins are locked in position in the bushing, and wear in service is minimal. The increased clearances on the occurrence aircraft, beyond overhaul manual limits, would have reduced stability. It was reported that a former operator had used a non-factory-approved O/H procedure for a period of about two years, in which the bushings were reamed for a loose fit on the pins. Both upper and lower torque links (T/L)s and pins were examined at the TSB Engineering Branch, and all materials and dimensions met the manufacturer's specifications. It was determined that the upper link failed in overload, with no evidence of pre-existing damage. The dimensions of both pins were equal to new parts limits, with no wear evident. The Engineering Branch report further states: There appears little doubt that the problem relates to the overall landing gear design and an inability to tolerate looseness or slack in the component build-up. The left oleo pressure and fluid had been discharged following the accident. The nitrogen pressure in the right oleo was 278 pounds per square inch (psi), and the fluid quantity was 10.5 litres (L). Servicing of both main oleos had taken place four days prior to the accident, and it was reported that the two oleo extensions had been equal prior to the flight. Specified nitrogen pressure in the oleos is 195/215 psi, and fluid quantity should be 13.1 L. This accident is the 29th occurrence of this type recorded in the aircraft manufacturer's data base. Numerous investigations and considerable research have been conducted over the previous 20 years and 28 occurrences, with the causes of landing gear shimmy emerging as a combination of any or all of the factors listed below (with comparison to this occurrence): The aircraft had been in operation with the regional airline for about 11 months, and had previously been operated in the UnitedStates. Since O/H in 1987 by the former operator, the landing gear had accumulated about 7,157 cycles. The operator's mid-life inspection of the gear was due at 8,000 cycles, with O/H required at 12,000 cycles. The T/L Apex Joint End Float inspection had been completed three days prior to the accident, and the clearance reset to specifications. The aircraft manufacturer had previously decided that the installation of a Torque Link Apex Damper (TLAD) would be the most effective means of eliminating or controlling the shimmy. Dutch Airworthiness Authorities (RLD) had proposed to require the installation of the TLAD kit on all Dutch-registered F28-Mk1000 and -Mk2000 series aircraft before 31December1993 by means of an Airworthiness Directive (AD) based on Fokker Service Bulletin (SB) F28/32-151. SB F28/32-151 was based on the original Dowty SB 32-169R. Because of problems with the TLAD development program, SB F28/32-151 was not issued, resulting in the compliance date of the proposed AD being postponed indefinitely. The aircraft manufacturer had previously taken the position that the probability of a T/L failure was relatively low, and the consequences minimal. However, because of the continuing failures of torque links and the recent incidences of substantial secondary damage to the aircraft involved, finding a solution to the shimmy problem has received higher priority. The regulatory authorities of Canada, the United States, and the Netherlands are monitoring further developments.